The Definitive Index
The AI1000 Rankings
The quarterly benchmark scoring the Top 1000 retailers by online sales on readiness for agentic commerce, co-developed by ReFiBuy and Digital Commerce 360. Sales rank shows who won the last era of ecommerce. The AI1000 shows who is positioned to win the next one.
The Q2 top five by AI1000 score — all five rank far down the online sales list. See who they are.
AI1000 rankings reshuffle in Q2, but retail challengers hold the lead.
-
Retail challengers held the lead, even as No. 1 changed hands.
Nixon, No. 722 by online sales, took No. 1, and all five Q2 leaders rank outside the 100 largest online retailers.
See the leaderboard → -
AI traffic leadership is spreading beyond ChatGPT.
ChatGPT was the largest AI traffic source for 844 retailers in Q1 and 722 in Q2, as Gemini, Perplexity, and Claude gained ground.
See the engine data → -
The readiness bar is rising faster than retailers are clearing it.
The AI1000 Index Average was 39.7 in Q2, compared with 42.0 in Q1, reflecting a more demanding environment as traffic spreads across more engines and selling surfaces.
972 of 1,000 retailers changed position in Q2.
Of the 1,000 retailers ranked, 972 changed position in Q2. The median retailer moved 35 spots, and 641 moved at least 25 positions. Traditional ecommerce leadership compounds over years. In AI shopping, nearly the entire board reshuffled in a single quarter.
Retailers by size of rank move between the Q1 and Q2 editions, in positions moved. 472 retailers rose and 500 fell; the median retailer moved 35 positions.
Retail challengers held the lead, even as No. 1 changed hands.
Nixon, a watch and accessories brand ranked No. 722 by online sales, overtook Online Labels, ranked No. 814, by 0.03 points. Four of Q1's top five remained in the top five, and all five Q2 leaders rank outside the 100 largest online retailers. Fast-moving retailers are jumping the line ahead of larger brands that have been slower to make their product data accessible to AI shopping agents.
| Q2 Rank | Movement | Retailer | AI1000 Score | Online Sales Rank |
|---|---|---|---|---|
| 1 | ▲ from 2 | Nixon | 69.72 | #722 |
| 2 | ▼ from 1 | Online Labels | 69.69 | #814 |
| 3 | ▲ from 4 | Everlane | 68.50 | #264 |
| 4 | ▼ from 3 | Fashionphile | 65.86 | #826 |
| 5 | ▲ from 9 | CustomInk | 64.69 | #133 |
No retailer moved into or out of the top 25 this quarter; the same 25 traded positions while the rest of the field moved around them. Eight of the Q1 top 10 remained in the Q2 top 10. Scores shown to two decimals.
Categories are moving at different speeds.
Rank movement varied sharply by category. Automotive Parts & Accessories climbed an average of 20.1 positions, while Flowers & Gifts fell 17.8, a spread of nearly 38 positions. The project-and-parts side of retail is climbing the rankings faster than the lifestyle aisles.
Average rank positions gained or lost per retailer within each category, Q1 to Q2. Positive means the category's retailers climbed the AI1000.
Automotive Parts & Accessories +20.1
The quarter's fastest climb. Parts retailers gained 20 positions on average, with the parts subcategory itself up 16.5.
Hardware & Home Improvement +12.1
Lawn and garden (+22.2), general home improvement (+18.6), and tools (+15.6) are among the fastest-climbing subcategories in the index.
Flowers & Gifts −17.8
The steepest slide of the quarter. Food & Beverage, Pet Care, and Health & Beauty also moved lower.
Category leaders held
Fourteen of the 15 category leaders remained at the top of their categories, and six rank outside the sales top 500.
Subcategory swings are sharper than the category averages.
At the subcategory level, the split is even clearer: lawn and garden, tools, and fishing and hunting retailers posted some of the quarter's biggest climbs, while beer, wine and spirits, books, and musical instrument sellers were among its largest decliners.
The eight biggest average climbs and the eight biggest slides among the 35 subcategories, Q1 to Q2; retailer counts per subcategory in parentheses.
AI traffic leadership is spreading beyond ChatGPT.
In Q1, ChatGPT was the largest source of AI-referred traffic for 844 of the 1,000 retailers. In Q2, that number fell to 722, while Gemini-led retailers doubled, Perplexity-led grew from eight to 21, and Claude-led increased from one to 15. A catalog tuned for one engine now has to perform across a growing number of AI surfaces.
Count of retailers whose largest AI traffic source is each engine, Q1 vs Q2 editions. The two panels use different scales so challenger movement stays legible. Counts of retailers, not traffic shares.
225 of the 1,000 largest retailers are detectably agent-ready.
Only 225 of the Top 1000 retailers have a detectable Universal Commerce Protocol (UCP) endpoint, part of an emerging standard that allows AI agents to connect directly with merchant catalogs. The retailers that do rank markedly higher in the AI1000. The catalogs AI agents will use this holiday are being prepared now.
Median AI1000 rank of the 225 retailers with a detectable UCP endpoint versus the rest, Q2 edition. Association, not attribution: UCP alone does not cause the rank difference.
New in Q2
Is your company on the AI1000?
Search the index to open your company's report: current position, quarter-over-quarter movement, and the signals behind your score.
Reports are private to each company and unlock by work-email verification · Not in the Top 1000? Run an Agentic Readiness Report
How the AI1000 is measured.
Every retailer's catalog is scored 0 to 100 on four signals measured from what AI shopping agents actually encounter: bot friendliness (including detectable UCP support), AI source traffic, diversity of AI sources, and 90-day momentum. The four combine into a single score and rank, recalculated every quarter. Scores reflect what agents can observe, not internal roadmaps.
The AI1000 is published quarterly. Its methodology is refined as measurement across AI shopping surfaces matures; rankings and index values reflect the methodology published for each edition.
Editions: Q2 2026 (current, this page) · Q1 2026, the first edition, published July 2026 · Q3 2026 is next.
“As AI-driven shopping becomes a more meaningful part of product discovery, retailers need new ways to understand how they are positioned. ReFiBuy brought the methodology and data capabilities to help us add an important new layer of agentic readiness to the Top 1000.”
Brian Warmoth, Editor-in-Chief, Digital Commerce 360
Don't miss the next edition.
AI shopping moves fast. Subscribe to get each AI1000 edition the morning it publishes.
Frequently asked questions
What is the AI1000?
The AI1000, published as the AI Commerce Rankings, is a quarterly benchmark that measures how ready the 1,000 largest online retailers are for AI shopping. It was co-developed by ReFiBuy and Digital Commerce 360 and scores every retailer in the Digital Commerce 360 Top 1000 from 0 to 100.
Who publishes the AI1000?
ReFiBuy and Digital Commerce 360 co-developed the index. ReFiBuy produces the underlying data and scoring; Digital Commerce 360 brings the Top 1000 retailer universe. New editions publish quarterly at refibuy.ai/ai1000.
How is the readiness score calculated?
Each retailer is scored on four signals measured from what AI shopping agents actually encounter: bot friendliness (whether agents can reach and read the catalog, including detectable UCP support), AI source traffic, diversity of AI sources, and 90-day momentum. The four combine into a single 0 to 100 score, recalculated every quarter.
What changed between Q1 and Q2 2026?
972 of the 1,000 retailers changed rank, and the median retailer moved 35 positions. Nixon, No. 722 by online sales, took No. 1 from Online Labels; AI traffic leadership spread beyond ChatGPT; and the AI1000 Index Average was 39.7, compared with 42.0 in Q1.
Which retailers rank highest in the Q2 2026 edition?
The Q2 top five are Nixon, Online Labels, Everlane, Fashionphile, and CustomInk. Three of the five rank outside the ecommerce sales top 700, and only 10 of the 100 largest online sellers place in the AI1000 top 100.
How often do the rankings update?
Quarterly. The Q2 2026 edition is current, and the Q3 edition is next. The methodology is refined as measurement across AI shopping surfaces matures; rankings and index values reflect the methodology published for each edition.
How does a retailer find its own rank?
Search for your company on this page to open its AI1000 report, then verify with a work email to unlock it. Retailers not in the Top 1000 can run ReFiBuy's Agentic Readiness Report for the same measurement.